# -*- coding: utf-8 -*-
import os
import sys
import re
import json
import time
import shutil
import warnings
import platform
import traceback
from subprocess import Popen
from multiprocessing import cpu_count

# -------------------------
# Version / Language
# -------------------------
if len(sys.argv) == 1:
    sys.argv.append("v2")
version = "v1" if sys.argv[1] == "v1" else "v2"
os.environ["version"] = version

now_dir = os.getcwd()
sys.path.insert(0, now_dir)
warnings.filterwarnings("ignore")

# -------------------------
# TEMP: per-session temp + cleanup old sessions
# -------------------------
tmp_root = os.path.join(now_dir, "TEMP")
os.makedirs(tmp_root, exist_ok=True)

session_tmp = os.path.join(tmp_root, str(int(time.time())))
os.makedirs(session_tmp, exist_ok=True)
os.environ["TEMP"] = session_tmp

# cleanup old TEMP sessions (3 days)
_now = time.time()
for name in os.listdir(tmp_root):
    path = os.path.join(tmp_root, name)
    if not os.path.isdir(path):
        continue
    try:
        ts = int(name)
        if _now - ts > 3 * 24 * 3600:
            shutil.rmtree(path, ignore_errors=True)
    except:
        # ignore non-timestamp folders
        pass

# -------------------------
# Safer path injection (no users.pth writing)
# -------------------------
extra_paths = [
    now_dir,
    os.path.join(now_dir, "tools"),
    os.path.join(now_dir, "tools", "asr"),
    os.path.join(now_dir, "GPT_SoVITS"),
    os.path.join(now_dir, "tools", "uvr5"),
]
for p in extra_paths:
    if os.path.isdir(p) and p not in sys.path:
        sys.path.insert(0, p)

os.environ["no_proxy"] = "localhost, 127.0.0.1, ::1"
os.environ["all_proxy"] = ""

# -------------------------
# Local imports (lightweight)
# -------------------------
from tools import my_utils
from tools.i18n.i18n import I18nAuto, scan_language_list
from config import (
    python_exec,
    infer_device,
    is_half,
    exp_root,
    webui_port_main,
    webui_port_infer_tts,
    webui_port_uvr5,
    webui_port_subfix,
    is_share,
)
from tools.my_utils import load_audio, check_for_existance, check_details
from tools.asr.config import asr_dict

language = sys.argv[-1] if sys.argv[-1] in scan_language_list() else "Auto"
os.environ["language"] = language
i18n = I18nAuto(language=language)

# Optional: gradio analytics version check disable (lazy)
def _disable_gradio_analytics_version_check():
    try:
        import gradio.analytics as analytics
        analytics.version_check = lambda: None
    except Exception:
        pass


# -------------------------
# Torch/GPU lazy helpers
# -------------------------
def _load_torch():
    import torch
    torch.manual_seed(233333)
    return torch


def _load_psutil():
    import psutil
    return psutil


def _gpu_probe():
    """
    More reliable GPU selection: CUDA available + VRAM >= 6GB
    Returns:
      gpu_info_str, gpus_str, default_gpu_number, default_batch_size, set_gpu_numbers
    """
    torch = _load_torch()
    psutil = _load_psutil()

    ngpu = torch.cuda.device_count()
    gpu_infos = []
    mem_gb = []
    set_gpu_numbers = set()

    if torch.cuda.is_available() and ngpu > 0:
        for i in range(ngpu):
            props = torch.cuda.get_device_properties(i)
            vram = props.total_memory / 1024**3
            name = props.name
            # 최소 6GB 이상이면 "훈련/가속 가능한 GPU"로 취급
            if vram >= 6:
                gpu_infos.append(f"{i}\t{name}")
                set_gpu_numbers.add(i)
                mem_gb.append(int(vram + 0.4))

    if gpu_infos:
        gpu_info_str = "\n".join(gpu_infos)
        default_batch_size = min(mem_gb) // 2
        gpus_str = "-".join([s.split("\t")[0] for s in gpu_infos])
        default_gpu_number = str(sorted(list(set_gpu_numbers))[0])
    else:
        gpu_info_str = "0\tCPU"
        gpus_str = "0"
        default_gpu_number = "0"
        set_gpu_numbers = {0}
        default_batch_size = int(psutil.virtual_memory().total / 1024 / 1024 / 1024 / 2)

    return gpu_info_str, gpus_str, default_gpu_number, default_batch_size, set_gpu_numbers


# -------------------------
# Process kill helpers
# -------------------------
SYSTEM = platform.system()

def kill_proc_tree(pid: int):
    """
    Cross-platform-ish process tree killer using psutil if available; fallback to taskkill on Windows.
    """
    if pid is None:
        return
    if SYSTEM == "Windows":
        os.system(f"taskkill /t /f /pid {pid}")
        return

    # non-windows
    try:
        psutil = _load_psutil()
        parent = psutil.Process(pid)
        for child in parent.children(recursive=True):
            try:
                child.terminate()
            except Exception:
                pass
        try:
            parent.terminate()
        except Exception:
            pass
    except Exception:
        # last resort
        try:
            os.kill(pid, 15)
        except Exception:
            pass


# -------------------------
# Weights discovery
# -------------------------
pretrained_sovits_name = [
    "GPT_SoVITS/pretrained_models/gsv-v2final-pretrained/s2G2333k.pth",
    "GPT_SoVITS/pretrained_models/s2G488k.pth",
]
pretrained_gpt_name = [
    "GPT_SoVITS/pretrained_models/gsv-v2final-pretrained/s1bert25hz-5kh-longer-epoch=12-step=369668.ckpt",
    "GPT_SoVITS/pretrained_models/s1bert25hz-2kh-longer-epoch=68e-step=50232.ckpt",
]

pretrained_model_list = (
    pretrained_sovits_name[-int(version[-1]) + 2],
    pretrained_sovits_name[-int(version[-1]) + 2].replace("s2G", "s2D"),
    pretrained_gpt_name[-int(version[-1]) + 2],
    "GPT_SoVITS/pretrained_models/chinese-roberta-wwm-ext-large",
    "GPT_SoVITS/pretrained_models/chinese-hubert-base",
)

_missing = ""
for p in pretrained_model_list:
    if not os.path.exists(p):
        _missing += f"\n    {p}"
if _missing:
    print("warning:", i18n("以下模型不存在:") + _missing)

_tmp = [[], []]
for i in range(2):
    _tmp[0].append(pretrained_gpt_name[i] if os.path.exists(pretrained_gpt_name[i]) else "")
    _tmp[1].append(pretrained_sovits_name[i] if os.path.exists(pretrained_sovits_name[i]) else "")
pretrained_gpt_name, pretrained_sovits_name = _tmp

SoVITS_weight_root = ["SoVITS_weights_v2", "SoVITS_weights"]
GPT_weight_root = ["GPT_weights_v2", "GPT_weights"]
for root in SoVITS_weight_root + GPT_weight_root:
    os.makedirs(root, exist_ok=True)


def get_weights_names():
    SoVITS_names = [name for name in pretrained_sovits_name if name]
    for path in SoVITS_weight_root:
        if os.path.isdir(path):
            for name in os.listdir(path):
                if name.endswith(".pth"):
                    SoVITS_names.append(f"{path}/{name}")

    GPT_names = [name for name in pretrained_gpt_name if name]
    for path in GPT_weight_root:
        if os.path.isdir(path):
            for name in os.listdir(path):
                if name.endswith(".ckpt"):
                    GPT_names.append(f"{path}/{name}")

    return SoVITS_names, GPT_names


def custom_sort_key(s: str):
    parts = re.split(r"(\d+)", s)
    return [int(x) if x.isdigit() else x for x in parts]


def change_choices():
    SoVITS_names, GPT_names = get_weights_names()
    return (
        {"choices": sorted(SoVITS_names, key=custom_sort_key), "__type__": "update"},
        {"choices": sorted(GPT_names, key=custom_sort_key), "__type__": "update"},
    )


# -------------------------
# GPU number sanitizers
# -------------------------
GPU_INFO_STR, GPUS_STR, DEFAULT_GPU_NUMBER, DEFAULT_BATCH_SIZE, SET_GPU_NUMBERS = _gpu_probe()

def fix_gpu_number(x: str):
    try:
        v = int(x)
        if v not in SET_GPU_NUMBERS:
            return DEFAULT_GPU_NUMBER
        return str(v)
    except Exception:
        return x

def fix_gpu_numbers(csv: str):
    try:
        items = []
        for t in csv.split(","):
            items.append(fix_gpu_number(t.strip()))
        return ",".join(items)
    except Exception:
        return csv


# -------------------------
# Subprocess handles
# -------------------------
p_label = None
p_uvr5 = None
p_asr = None
p_denoise = None
p_tts_inference = None
p_train_sovits = None
p_train_gpt = None
ps_slice = []
ps1a = []
ps1b = []
ps1c = []
ps1abc = []


# -------------------------
# Tools: label / uvr5 / tts inference
# -------------------------
def change_label(path_list):
    global p_label
    import gradio as gr

    if p_label is None:
        check_for_existance([path_list])
        path_list = my_utils.clean_path(path_list)

        cmd = [
            python_exec,
            "tools/subfix_webui.py",
            "--load_list",
            path_list,
            "--webui_port",
            str(webui_port_subfix),
            "--is_share",
            str(is_share),
        ]
        yield i18n("打标工具WebUI已开启"), {"__type__": "update", "visible": False}, {"__type__": "update", "visible": True}
        print(" ".join(cmd))
        p_label = Popen(cmd)
    else:
        kill_proc_tree(p_label.pid)
        p_label = None
        yield i18n("打标工具WebUI已关闭"), {"__type__": "update", "visible": True}, {"__type__": "update", "visible": False}


def change_uvr5():
    global p_uvr5

    if p_uvr5 is None:
        cmd = [
            python_exec,
            "tools/uvr5/webui.py",
            str(infer_device),
            str(is_half),
            str(webui_port_uvr5),
            str(is_share),
        ]
        yield i18n("UVR5已开启"), {"__type__": "update", "visible": False}, {"__type__": "update", "visible": True}
        print(" ".join(cmd))
        p_uvr5 = Popen(cmd)
    else:
        kill_proc_tree(p_uvr5.pid)
        p_uvr5 = None
        yield i18n("UVR5已关闭"), {"__type__": "update", "visible": True}, {"__type__": "update", "visible": False}


def change_tts_inference(bert_path, cnhubert_base_path, gpu_number, gpt_path, sovits_path, batched_infer_enabled):
    global p_tts_inference

    if p_tts_inference is None:
        os.environ["gpt_path"] = gpt_path if "/" in gpt_path else f"{GPT_weight_root}/{gpt_path}"
        os.environ["sovits_path"] = sovits_path if "/" in sovits_path else f"{SoVITS_weight_root}/{sovits_path}"
        os.environ["cnhubert_base_path"] = cnhubert_base_path
        os.environ["bert_path"] = bert_path
        os.environ["_CUDA_VISIBLE_DEVICES"] = fix_gpu_number(gpu_number)
        os.environ["is_half"] = str(is_half)
        os.environ["infer_ttswebui"] = str(webui_port_infer_tts)
        os.environ["is_share"] = str(is_share)

        if batched_infer_enabled:
            cmd = [python_exec, "GPT_SoVITS/inference_webui_fast.py", str(language)]
        else:
            cmd = [python_exec, "GPT_SoVITS/inference_webui.py", str(language)]

        yield i18n("TTS推理进程已开启"), {"__type__": "update", "visible": False}, {"__type__": "update", "visible": True}
        print(" ".join(cmd))
        p_tts_inference = Popen(cmd)
    else:
        kill_proc_tree(p_tts_inference.pid)
        p_tts_inference = None
        yield i18n("TTS推理进程已关闭"), {"__type__": "update", "visible": True}, {"__type__": "update", "visible": False}


# -------------------------
# ASR / denoise / slicer
# -------------------------
def open_asr(asr_inp_dir, asr_opt_dir, asr_model, asr_model_size, asr_lang, asr_precision):
    global p_asr
    if p_asr is not None:
        yield "已有正在进行的ASR任务，需先终止才能开启下一次任务", {"__type__":"update","visible":False}, {"__type__":"update","visible":True}, {"__type__":"update"}, {"__type__":"update"}, {"__type__":"update"}
        return

    asr_inp_dir = my_utils.clean_path(asr_inp_dir)
    asr_opt_dir = my_utils.clean_path(asr_opt_dir)
    check_for_existance([asr_inp_dir])

    cmd = [python_exec, f"tools/asr/{asr_dict[asr_model]['path']}"]
    cmd += ["-i", asr_inp_dir]
    cmd += ["-o", asr_opt_dir]
    cmd += ["-s", str(asr_model_size)]
    cmd += ["-l", str(asr_lang)]
    cmd += ["-p", str(asr_precision)]

    output_file_name = os.path.basename(asr_inp_dir)
    output_folder = asr_opt_dir or "output/asr_opt"
    output_file_path = os.path.abspath(f"{output_folder}/{output_file_name}.list")

    yield f"ASR任务开启：{' '.join(cmd)}", {"__type__":"update","visible":False}, {"__type__":"update","visible":True}, {"__type__":"update"}, {"__type__":"update"}, {"__type__":"update"}
    print(" ".join(cmd))
    p_asr = Popen(cmd)
    p_asr.wait()
    p_asr = None

    yield (
        "ASR任务完成, 查看终端进行下一步",
        {"__type__":"update","visible":True},
        {"__type__":"update","visible":False},
        {"__type__":"update","value":output_file_path},
        {"__type__":"update","value":output_file_path},
        {"__type__":"update","value":asr_inp_dir},
    )

def close_asr():
    global p_asr
    if p_asr is not None:
        kill_proc_tree(p_asr.pid)
        p_asr = None
    return "已终止ASR进程", {"__type__":"update","visible":True}, {"__type__":"update","visible":False}

def open_denoise(denoise_inp_dir, denoise_opt_dir):
    global p_denoise
    if p_denoise is not None:
        yield "已有正在进行的语音降噪任务，需先终止才能开启下一次任务", {"__type__":"update","visible":False}, {"__type__":"update","visible":True}, {"__type__":"update"}, {"__type__":"update"}
        return

    denoise_inp_dir = my_utils.clean_path(denoise_inp_dir)
    denoise_opt_dir = my_utils.clean_path(denoise_opt_dir)
    check_for_existance([denoise_inp_dir])

    precision = "float16" if is_half else "float32"
    cmd = [python_exec, "tools/cmd-denoise.py", "-i", denoise_inp_dir, "-o", denoise_opt_dir, "-p", precision]

    yield f"语音降噪任务开启：{' '.join(cmd)}", {"__type__":"update","visible":False}, {"__type__":"update","visible":True}, {"__type__":"update"}, {"__type__":"update"}
    print(" ".join(cmd))
    p_denoise = Popen(cmd)
    p_denoise.wait()
    p_denoise = None

    yield (
        "语音降噪任务完成, 查看终端进行下一步",
        {"__type__":"update","visible":True},
        {"__type__":"update","visible":False},
        {"__type__":"update","value":denoise_opt_dir},
        {"__type__":"update","value":denoise_opt_dir},
    )

def close_denoise():
    global p_denoise
    if p_denoise is not None:
        kill_proc_tree(p_denoise.pid)
        p_denoise = None
    return "已终止语音降噪进程", {"__type__":"update","visible":True}, {"__type__":"update","visible":False}

def open_slice(inp, opt_root, threshold, min_length, min_interval, hop_size, max_sil_kept, _max, alpha, n_parts):
    global ps_slice
    inp = my_utils.clean_path(inp)
    opt_root = my_utils.clean_path(opt_root)
    check_for_existance([inp])

    if not os.path.exists(inp):
        yield "输入路径不存在", {"__type__":"update","visible":True}, {"__type__":"update","visible":False}, {"__type__":"update"}, {"__type__":"update"}, {"__type__":"update"}
        return

    if os.path.isfile(inp):
        n_parts = 1
    elif os.path.isdir(inp):
        pass
    else:
        yield "输入路径存在但既不是文件也不是文件夹", {"__type__":"update","visible":True}, {"__type__":"update","visible":False}, {"__type__":"update"}, {"__type__":"update"}, {"__type__":"update"}
        return

    if ps_slice:
        yield "已有正在进行的切割任务，需先终止才能开启下一次任务", {"__type__":"update","visible":False}, {"__type__":"update","visible":True}, {"__type__":"update"}, {"__type__":"update"}, {"__type__":"update"}
        return

    for i_part in range(int(n_parts)):
        cmd = [
            python_exec,
            "tools/slice_audio.py",
            inp,
            opt_root,
            str(threshold),
            str(min_length),
            str(min_interval),
            str(hop_size),
            str(max_sil_kept),
            str(_max),
            str(alpha),
            str(i_part),
            str(n_parts),
        ]
        print(" ".join(cmd))
        ps_slice.append(Popen(cmd))

    yield "切割执行中", {"__type__":"update","visible":False}, {"__type__":"update","visible":True}, {"__type__":"update"}, {"__type__":"update"}, {"__type__":"update"}

    for p in ps_slice:
        p.wait()

    ps_slice = []
    yield (
        "切割结束",
        {"__type__":"update","visible":True},
        {"__type__":"update","visible":False},
        {"__type__":"update","value":opt_root},
        {"__type__":"update","value":opt_root},
        {"__type__":"update","value":opt_root},
    )

def close_slice():
    global ps_slice
    if ps_slice:
        for p in ps_slice:
            try:
                kill_proc_tree(p.pid)
            except Exception:
                traceback.print_exc()
        ps_slice = []
    return "已终止所有切割进程", {"__type__":"update","visible":True}, {"__type__":"update","visible":False}


# -------------------------
# Dataset prep (1a/1b/1c/1abc)
# -------------------------
def open1a(inp_text, inp_wav_dir, exp_name, gpu_numbers_text, bert_pretrained_dir):
    global ps1a
    inp_text = my_utils.clean_path(inp_text)
    inp_wav_dir = my_utils.clean_path(inp_wav_dir)

    if check_for_existance([inp_text, inp_wav_dir], is_dataset_processing=True):
        check_details([inp_text, inp_wav_dir], is_dataset_processing=True)

    if ps1a:
        yield "已有正在进行的文本任务，需先终止才能开启下一次任务", {"__type__":"update","visible":False}, {"__type__":"update","visible":True}
        return

    opt_dir = f"{exp_root}/{exp_name}"
    config = {
        "inp_text": inp_text,
        "inp_wav_dir": inp_wav_dir,
        "exp_name": exp_name,
        "opt_dir": opt_dir,
        "bert_pretrained_dir": bert_pretrained_dir,
        "is_half": str(is_half),
    }

    gpu_names = gpu_numbers_text.split("-")
    all_parts = len(gpu_names)

    for i_part in range(all_parts):
        cfg = dict(config)
        cfg.update(
            {
                "i_part": str(i_part),
                "all_parts": str(all_parts),
                "_CUDA_VISIBLE_DEVICES": fix_gpu_number(gpu_names[i_part]),
            }
        )
        os.environ.update(cfg)
        cmd = [python_exec, "GPT_SoVITS/prepare_datasets/1-get-text.py"]
        print(" ".join(cmd))
        ps1a.append(Popen(cmd))

    yield "文本进程执行中", {"__type__":"update","visible":False}, {"__type__":"update","visible":True}
    for p in ps1a:
        p.wait()

    # merge
    opt = []
    for i_part in range(all_parts):
        txt_path = f"{opt_dir}/2-name2text-{i_part}.txt"
        if os.path.exists(txt_path):
            with open(txt_path, "r", encoding="utf8") as f:
                opt += f.read().strip("\n").split("\n")
            try:
                os.remove(txt_path)
            except:
                pass

    path_text = f"{opt_dir}/2-name2text.txt"
    os.makedirs(opt_dir, exist_ok=True)
    with open(path_text, "w", encoding="utf8") as f:
        f.write("\n".join([x for x in opt if x.strip()]) + "\n")

    ps1a = []
    if len("".join(opt)) > 0:
        yield "文本进程成功", {"__type__":"update","visible":True}, {"__type__":"update","visible":False}
    else:
        yield "文本进程失败", {"__type__":"update","visible":True}, {"__type__":"update","visible":False}

def close1a():
    global ps1a
    if ps1a:
        for p in ps1a:
            try:
                kill_proc_tree(p.pid)
            except:
                traceback.print_exc()
        ps1a = []
    return "已终止所有1a进程", {"__type__":"update","visible":True}, {"__type__":"update","visible":False}

def open1b(inp_text, inp_wav_dir, exp_name, gpu_numbers_ssl, ssl_pretrained_dir):
    global ps1b
    inp_text = my_utils.clean_path(inp_text)
    inp_wav_dir = my_utils.clean_path(inp_wav_dir)

    if check_for_existance([inp_text, inp_wav_dir], is_dataset_processing=True):
        check_details([inp_text, inp_wav_dir], is_dataset_processing=True)

    if ps1b:
        yield "已有正在进行的SSL提取任务，需先终止才能开启下一次任务", {"__type__":"update","visible":False}, {"__type__":"update","visible":True}
        return

    config = {
        "inp_text": inp_text,
        "inp_wav_dir": inp_wav_dir,
        "exp_name": exp_name,
        "opt_dir": f"{exp_root}/{exp_name}",
        "cnhubert_base_dir": ssl_pretrained_dir,
        "is_half": str(is_half),
    }

    gpu_names = gpu_numbers_ssl.split("-")
    all_parts = len(gpu_names)

    for i_part in range(all_parts):
        cfg = dict(config)
        cfg.update(
            {
                "i_part": str(i_part),
                "all_parts": str(all_parts),
                "_CUDA_VISIBLE_DEVICES": fix_gpu_number(gpu_names[i_part]),
            }
        )
        os.environ.update(cfg)
        cmd = [python_exec, "GPT_SoVITS/prepare_datasets/2-get-hubert-wav32k.py"]
        print(" ".join(cmd))
        ps1b.append(Popen(cmd))

    yield "SSL提取进程执行中", {"__type__":"update","visible":False}, {"__type__":"update","visible":True}
    for p in ps1b:
        p.wait()

    ps1b = []
    yield "SSL提取进程结束", {"__type__":"update","visible":True}, {"__type__":"update","visible":False}

def close1b():
    global ps1b
    if ps1b:
        for p in ps1b:
            try:
                kill_proc_tree(p.pid)
            except:
                traceback.print_exc()
        ps1b = []
    return "已终止所有1b进程", {"__type__":"update","visible":True}, {"__type__":"update","visible":False}

def open1c(inp_text, exp_name, gpu_numbers_sem, pretrained_s2G_path):
    global ps1c
    inp_text = my_utils.clean_path(inp_text)

    if check_for_existance([inp_text, ""], is_dataset_processing=True):
        check_details([inp_text, ""], is_dataset_processing=True)

    if ps1c:
        yield "已有正在进行的语义token提取任务，需先终止才能开启下一次任务", {"__type__":"update","visible":False}, {"__type__":"update","visible":True}
        return

    opt_dir = f"{exp_root}/{exp_name}"
    config = {
        "inp_text": inp_text,
        "exp_name": exp_name,
        "opt_dir": opt_dir,
        "pretrained_s2G": pretrained_s2G_path,
        "s2config_path": "GPT_SoVITS/configs/s2.json",
        "is_half": str(is_half),
    }

    gpu_names = gpu_numbers_sem.split("-")
    all_parts = len(gpu_names)

    for i_part in range(all_parts):
        cfg = dict(config)
        cfg.update(
            {
                "i_part": str(i_part),
                "all_parts": str(all_parts),
                "_CUDA_VISIBLE_DEVICES": fix_gpu_number(gpu_names[i_part]),
            }
        )
        os.environ.update(cfg)
        cmd = [python_exec, "GPT_SoVITS/prepare_datasets/3-get-semantic.py"]
        print(" ".join(cmd))
        ps1c.append(Popen(cmd))

    yield "语义token提取进程执行中", {"__type__":"update","visible":False}, {"__type__":"update","visible":True}
    for p in ps1c:
        p.wait()

    # merge
    os.makedirs(opt_dir, exist_ok=True)
    opt = ["item_name\tsemantic_audio"]
    path_semantic = f"{opt_dir}/6-name2semantic.tsv"

    for i_part in range(all_parts):
        semantic_path = f"{opt_dir}/6-name2semantic-{i_part}.tsv"
        if os.path.exists(semantic_path):
            with open(semantic_path, "r", encoding="utf8") as f:
                opt += f.read().strip("\n").split("\n")
            try:
                os.remove(semantic_path)
            except:
                pass

    with open(path_semantic, "w", encoding="utf8") as f:
        f.write("\n".join(opt) + "\n")

    ps1c = []
    yield "语义token提取进程结束", {"__type__":"update","visible":True}, {"__type__":"update","visible":False}

def close1c():
    global ps1c
    if ps1c:
        for p in ps1c:
            try:
                kill_proc_tree(p.pid)
            except:
                traceback.print_exc()
        ps1c = []
    return "已终止所有语义token进程", {"__type__":"update","visible":True}, {"__type__":"update","visible":False}


def open1abc(inp_text, inp_wav_dir, exp_name, gpu_numbers_text, gpu_numbers_ssl, gpu_numbers_sem, bert_pretrained_dir, ssl_pretrained_dir, pretrained_s2G_path):
    global ps1abc
    if ps1abc:
        yield "已有正在进行的一键三连任务，需先终止才能开启下一次任务", {"__type__":"update","visible":False}, {"__type__":"update","visible":True}
        return

    try:
        # 1a
        gen1a = open1a(inp_text, inp_wav_dir, exp_name, gpu_numbers_text, bert_pretrained_dir)
        for x in gen1a:
            yield x

        # 1b
        gen1b = open1b(inp_text, inp_wav_dir, exp_name, gpu_numbers_ssl, ssl_pretrained_dir)
        for x in gen1b:
            yield x

        # 1c
        gen1c = open1c(inp_text, exp_name, gpu_numbers_sem, pretrained_s2G_path)
        for x in gen1c:
            yield x

        yield "一键三连进程结束", {"__type__":"update","visible":True}, {"__type__":"update","visible":False}
    except Exception:
        traceback.print_exc()
        close1abc()
        yield "一键三连中途报错", {"__type__":"update","visible":True}, {"__type__":"update","visible":False}

def close1abc():
    global ps1abc
    # 이 버전은 open1abc가 내부적으로 open1a/1b/1c를 호출하므로,
    # 각각 close를 호출해도 되는데, 여기서는 안전하게 리스트를 비우고 안내만 한다.
    ps1abc = []
    return "已终止所有一键三连进程", {"__type__":"update","visible":True}, {"__type__":"update","visible":False}


# -------------------------
# Train (SoVITS / GPT)
# -------------------------
def open1Ba(batch_size, total_epoch, exp_name, text_low_lr_rate, if_save_latest, if_save_every_weights, save_every_epoch, gpu_numbers_sovits_train, pretrained_s2G, pretrained_s2D):
    global p_train_sovits
    if p_train_sovits is not None:
        yield "已有正在进行的SoVITS训练任务，需先终止才能开启下一次任务", {"__type__":"update","visible":False}, {"__type__":"update","visible":True}
        return

    import yaml  # optional; for compatibility (not used here, but keep)
    torch = _load_torch()

    with open("GPT_SoVITS/configs/s2.json", "r", encoding="utf8") as f:
        data = json.loads(f.read())

    s2_dir = f"{exp_root}/{exp_name}"
    os.makedirs(f"{s2_dir}/logs_s2", exist_ok=True)
    if check_for_existance([s2_dir], is_train=True):
        check_details([s2_dir], is_train=True)

    if is_half is False:
        data["train"]["fp16_run"] = False
        batch_size = max(1, int(batch_size) // 2)

    data["train"]["batch_size"] = int(batch_size)
    data["train"]["epochs"] = int(total_epoch)
    data["train"]["text_low_lr_rate"] = float(text_low_lr_rate)
    data["train"]["pretrained_s2G"] = pretrained_s2G
    data["train"]["pretrained_s2D"] = pretrained_s2D
    data["train"]["if_save_latest"] = bool(if_save_latest)
    data["train"]["if_save_every_weights"] = bool(if_save_every_weights)
    data["train"]["save_every_epoch"] = int(save_every_epoch)
    data["train"]["gpu_numbers"] = gpu_numbers_sovits_train
    data["model"]["version"] = version
    data["data"]["exp_dir"] = data["s2_ckpt_dir"] = s2_dir
    data["save_weight_dir"] = SoVITS_weight_root[-int(version[-1]) + 2]
    data["name"] = exp_name
    data["version"] = version

    tmp_config_path = os.path.join(session_tmp, "tmp_s2.json")
    with open(tmp_config_path, "w", encoding="utf8") as f:
        f.write(json.dumps(data, ensure_ascii=False))

    cmd = [python_exec, "GPT_SoVITS/s2_train.py", "--config", tmp_config_path]
    yield f"SoVITS训练开始：{' '.join(cmd)}", {"__type__":"update","visible":False}, {"__type__":"update","visible":True}
    print(" ".join(cmd))
    p_train_sovits = Popen(cmd)
    p_train_sovits.wait()
    p_train_sovits = None
    yield "SoVITS训练完成", {"__type__":"update","visible":True}, {"__type__":"update","visible":False}

def close1Ba():
    global p_train_sovits
    if p_train_sovits is not None:
        kill_proc_tree(p_train_sovits.pid)
        p_train_sovits = None
    return "已终止SoVITS训练", {"__type__":"update","visible":True}, {"__type__":"update","visible":False}


def open1Bb(batch_size, total_epoch, exp_name, if_dpo, if_save_latest, if_save_every_weights, save_every_epoch, gpu_numbers_gpt_train, pretrained_s1):
    global p_train_gpt
    if p_train_gpt is not None:
        yield "已有正在进行的GPT训练任务，需先终止才能开启下一次任务", {"__type__":"update","visible":False}, {"__type__":"update","visible":True}
        return

    import yaml

    cfg_path = "GPT_SoVITS/configs/s1longer.yaml" if version == "v1" else "GPT_SoVITS/configs/s1longer-v2.yaml"
    with open(cfg_path, "r", encoding="utf8") as f:
        data = yaml.load(f.read(), Loader=yaml.FullLoader)

    s1_dir = f"{exp_root}/{exp_name}"
    os.makedirs(f"{s1_dir}/logs_s1", exist_ok=True)
    if check_for_existance([s1_dir], is_train=True):
        check_details([s1_dir], is_train=True)

    if is_half is False:
        data["train"]["precision"] = "32"
        batch_size = max(1, int(batch_size) // 2)

    data["train"]["batch_size"] = int(batch_size)
    data["train"]["epochs"] = int(total_epoch)
    data["pretrained_s1"] = pretrained_s1
    data["train"]["save_every_n_epoch"] = int(save_every_epoch)
    data["train"]["if_save_every_weights"] = bool(if_save_every_weights)
    data["train"]["if_save_latest"] = bool(if_save_latest)
    data["train"]["if_dpo"] = bool(if_dpo)
    data["train"]["half_weights_save_dir"] = GPT_weight_root[-int(version[-1]) + 2]
    data["train"]["exp_name"] = exp_name
    data["train_semantic_path"] = f"{s1_dir}/6-name2semantic.tsv"
    data["train_phoneme_path"] = f"{s1_dir}/2-name2text.txt"
    data["output_dir"] = f"{s1_dir}/logs_s1"

    os.environ["_CUDA_VISIBLE_DEVICES"] = fix_gpu_numbers(gpu_numbers_gpt_train.replace("-", ","))
    os.environ["hz"] = "25hz"

    tmp_config_path = os.path.join(session_tmp, "tmp_s1.yaml")
    with open(tmp_config_path, "w", encoding="utf8") as f:
        f.write(yaml.dump(data, default_flow_style=False, allow_unicode=True))

    cmd = [python_exec, "GPT_SoVITS/s1_train.py", "--config_file", tmp_config_path]
    yield f"GPT训练开始：{' '.join(cmd)}", {"__type__":"update","visible":False}, {"__type__":"update","visible":True}
    print(" ".join(cmd))
    p_train_gpt = Popen(cmd)
    p_train_gpt.wait()
    p_train_gpt = None
    yield "GPT训练完成", {"__type__":"update","visible":True}, {"__type__":"update","visible":False}

def close1Bb():
    global p_train_gpt
    if p_train_gpt is not None:
        kill_proc_tree(p_train_gpt.pid)
        p_train_gpt = None
    return "已终止GPT训练", {"__type__":"update","visible":True}, {"__type__":"update","visible":False}


# -------------------------
# Switch version
# -------------------------
def switch_version(version_):
    import gradio as gr
    os.environ["version"] = version_
    global version
    version = version_
    if not (pretrained_sovits_name[-int(version[-1]) + 2] and pretrained_gpt_name[-int(version[-1]) + 2]):
        gr.Warning(i18n(f"未下载{version.upper()}模型"))
    return (
        {"__type__":"update", "value": pretrained_sovits_name[-int(version[-1]) + 2]},
        {"__type__":"update", "value": pretrained_sovits_name[-int(version[-1]) + 2].replace("s2G","s2D")},
        {"__type__":"update", "value": pretrained_gpt_name[-int(version[-1]) + 2]},
        {"__type__":"update", "value": pretrained_gpt_name[-int(version[-1]) + 2]},
        {"__type__":"update", "value": pretrained_sovits_name[-int(version[-1]) + 2]},
    )


def sync(text):
    return {"__type__": "update", "value": text}


# -------------------------
# Ensure G2PWModel
# -------------------------
if os.path.exists("GPT_SoVITS/text/G2PWModel"):
    pass
else:
    cmd = [python_exec, "GPT_SoVITS/download.py"]
    p = Popen(cmd)
    p.wait()


# -------------------------
# Gradio UI
# -------------------------
def main():
    _disable_gradio_analytics_version_check()
    import gradio as gr

    n_cpu = cpu_count()

    with gr.Blocks(title="GPT-SoVITS WebUI") as app:
        gr.Markdown(value=i18n("本软件以MIT协议开源, 作者不对软件具备任何控制力, 使用软件者、传播软件导出的声音者自负全责. <br>如不认可该条款, 则不能使用或引用软件包内任何代码和文件. 详见根目录<b>LICENSE</b>."))
        gr.Markdown(value=i18n("中文教程文档：https://www.yuque.com/baicaigongchang1145haoyuangong/ib3g1e"))

        with gr.Tabs():
            # 0 - tools
            with gr.TabItem(i18n("0-前置数据集获取工具")):
                gr.Markdown(value=i18n("0a-UVR5人声伴奏分离&去混响去延迟工具"))
                with gr.Row():
                    with gr.Column(scale=3):
                        uvr5_info = gr.Textbox(label=i18n("UVR5进程输出信息"))
                    open_uvr5 = gr.Button(value=i18n("开启UVR5-WebUI"), variant="primary", visible=True)
                    close_uvr5 = gr.Button(value=i18n("关闭UVR5-WebUI"), variant="primary", visible=False)

                gr.Markdown(value=i18n("0b-语音切分工具"))
                with gr.Row():
                    with gr.Column(scale=3):
                        with gr.Row():
                            slice_inp_path = gr.Textbox(label=i18n("音频自动切分输入路径，可文件可文件夹"), value="")
                            slice_opt_root = gr.Textbox(label=i18n("切分后的子音频的输出根目录"), value="output/slicer_opt")
                        with gr.Row():
                            threshold = gr.Textbox(label=i18n("threshold:音量小于这个值视作静音的备选切割点"), value="-34")
                            min_length = gr.Textbox(label=i18n("min_length:每段最小多长，如果第一段太短一直和后面段连起来直到超过这个值"), value="4000")
                            min_interval = gr.Textbox(label=i18n("min_interval:最短切割间隔"), value="300")
                            hop_size = gr.Textbox(label=i18n("hop_size:怎么算音量曲线，越小精度越大计算量越高（不是精度越大效果越好）"), value="10")
                            max_sil_kept = gr.Textbox(label=i18n("max_sil_kept:切完后静音最多留多长"), value="500")
                        with gr.Row():
                            _max = gr.Slider(minimum=0, maximum=1, step=0.05, label=i18n("max:归一化后最大值多少"), value=0.9, interactive=True)
                            alpha = gr.Slider(minimum=0, maximum=1, step=0.05, label=i18n("alpha_mix:混多少比例归一化后音频进来"), value=0.25, interactive=True)
                        with gr.Row():
                            n_process = gr.Slider(minimum=1, maximum=n_cpu, step=1, label=i18n("切割使用的进程数"), value=4, interactive=True)
                            slicer_info = gr.Textbox(label=i18n("语音切割进程输出信息"))
                    open_slicer_button = gr.Button(i18n("开启语音切割"), variant="primary", visible=True)
                    close_slicer_button = gr.Button(i18n("终止语音切割"), variant="primary", visible=False)

                gr.Markdown(value=i18n("0bb-语音降噪工具"))
                with gr.Row():
                    with gr.Column(scale=3):
                        with gr.Row():
                            denoise_input_dir = gr.Textbox(label=i18n("降噪音频文件输入文件夹"), value="")
                            denoise_output_dir = gr.Textbox(label=i18n("降噪结果输出文件夹"), value="output/denoise_opt")
                        denoise_info = gr.Textbox(label=i18n("语音降噪进程输出信息"))
                    open_denoise_button = gr.Button(i18n("开启语音降噪"), variant="primary", visible=True)
                    close_denoise_button = gr.Button(i18n("终止语音降噪进程"), variant="primary", visible=False)

                gr.Markdown(value=i18n("0c-中文批量离线ASR工具"))
                with gr.Row():
                    with gr.Column(scale=3):
                        with gr.Row():
                            asr_inp_dir = gr.Textbox(label=i18n("输入文件夹路径"), value=r"D:\GPT-SoVITS\raw\xxx", interactive=True)
                            asr_opt_dir = gr.Textbox(label=i18n("输出文件夹路径"), value="output/asr_opt", interactive=True)
                        with gr.Row():
                            asr_model = gr.Dropdown(label=i18n("ASR 模型"), choices=list(asr_dict.keys()), interactive=True, value="达摩 ASR (中文)")
                            asr_size = gr.Dropdown(label=i18n("ASR 模型尺寸"), choices=["large"], interactive=True, value="large")
                            asr_lang = gr.Dropdown(label=i18n("ASR 语言设置"), choices=["zh", "yue"], interactive=True, value="zh")
                            asr_precision = gr.Dropdown(label=i18n("数据类型精度"), choices=["float32"], interactive=True, value="float32")
                        asr_info = gr.Textbox(label=i18n("ASR进程输出信息"))
                    open_asr_button = gr.Button(i18n("开启离线批量ASR"), variant="primary", visible=True)
                    close_asr_button = gr.Button(i18n("终止ASR进程"), variant="primary", visible=False)

                def change_lang_choices(key):
                    return {"__type__":"update","choices":asr_dict[key]["lang"],"value":asr_dict[key]["lang"][0]}
                def change_size_choices(key):
                    return {"__type__":"update","choices":asr_dict[key]["size"],"value":asr_dict[key]["size"][-1]}
                def change_precision_choices(key):
                    # Faster Whisper면 상황 따라 바꾸는 로직을 유지
                    if key == "Faster Whisper (多语种)":
                        if DEFAULT_BATCH_SIZE <= 4:
                            precision = "int8"
                        elif is_half:
                            precision = "float16"
                        else:
                            precision = "float32"
                    else:
                        precision = "float32"
                    return {"__type__":"update","choices":asr_dict[key]["precision"],"value":precision}

                asr_model.change(change_lang_choices, [asr_model], [asr_lang])
                asr_model.change(change_size_choices, [asr_model], [asr_size])
                asr_model.change(change_precision_choices, [asr_model], [asr_precision])

                gr.Markdown(value=i18n("0d-语音文本校对标注工具"))
                with gr.Row():
                    with gr.Column(scale=3):
                        path_list = gr.Textbox(label=i18n(".list标注文件的路径"), value=r"D:\RVC1006\GPT-SoVITS\raw\xxx.list", interactive=True)
                        label_info = gr.Textbox(label=i18n("打标工具进程输出信息"))
                    open_label = gr.Button(value=i18n("开启打标WebUI"), variant="primary", visible=True)
                    close_label = gr.Button(value=i18n("关闭打标WebUI"), variant="primary", visible=False)

                open_label.click(change_label, [path_list], [label_info, open_label, close_label])
                close_label.click(change_label, [path_list], [label_info, open_label, close_label])
                open_uvr5.click(change_uvr5, [], [uvr5_info, open_uvr5, close_uvr5])
                close_uvr5.click(change_uvr5, [], [uvr5_info, open_uvr5, close_uvr5])

                open_asr_button.click(open_asr, [asr_inp_dir, asr_opt_dir, asr_model, asr_size, asr_lang, asr_precision], [asr_info, open_asr_button, close_asr_button, path_list, path_list, denoise_input_dir])
                close_asr_button.click(close_asr, [], [asr_info, open_asr_button, close_asr_button])

                open_slicer_button.click(open_slice, [slice_inp_path, slice_opt_root, threshold, min_length, min_interval, hop_size, max_sil_kept, _max, alpha, n_process], [slicer_info, open_slicer_button, close_slicer_button, asr_inp_dir, denoise_input_dir, denoise_input_dir])
                close_slicer_button.click(close_slice, [], [slicer_info, open_slicer_button, close_slicer_button])

                open_denoise_button.click(open_denoise, [denoise_input_dir, denoise_output_dir], [denoise_info, open_denoise_button, close_denoise_button, asr_inp_dir, denoise_input_dir])
                close_denoise_button.click(close_denoise, [], [denoise_info, open_denoise_button, close_denoise_button])

            # 1 - TTS
            with gr.TabItem(i18n("1-GPT-SoVITS-TTS")):
                with gr.Row():
                    exp_name = gr.Textbox(label=i18n("*实验/模型名"), value="xxx", interactive=True)
                    gpu_info_box = gr.Textbox(label=i18n("显卡信息"), value=GPU_INFO_STR, visible=True, interactive=False)
                    version_checkbox = gr.Radio(label=i18n("版本"), value=version, choices=["v1", "v2"])

                with gr.Row():
                    pretrained_s2G = gr.Textbox(label=i18n("预训练的SoVITS-G模型路径"), value=pretrained_sovits_name[-int(version[-1]) + 2], interactive=True, lines=2, max_lines=3, scale=9)
                    pretrained_s2D = gr.Textbox(label=i18n("预训练的SoVITS-D模型路径"), value=pretrained_sovits_name[-int(version[-1]) + 2].replace("s2G", "s2D"), interactive=True, lines=2, max_lines=3, scale=9)
                    pretrained_s1 = gr.Textbox(label=i18n("预训练的GPT模型路径"), value=pretrained_gpt_name[-int(version[-1]) + 2], interactive=True, lines=2, max_lines=3, scale=10)

                with gr.TabItem(i18n("1A-训练集格式化工具")):
                    gr.Markdown(value=i18n("输出logs/实验名目录下应有23456开头的文件和文件夹"))
                    with gr.Row():
                        inp_text = gr.Textbox(label=i18n("*文本标注文件"), value=r"D:\RVC1006\GPT-SoVITS\raw\xxx.list", interactive=True, scale=10)
                        inp_wav_dir = gr.Textbox(
                            label=i18n("*训练集音频文件目录"),
                            interactive=True,
                            placeholder=i18n("填切割后音频所在目录！读取的音频文件完整路径=该目录-拼接-list文件里波形对应的文件名（不是全路径）。如果留空则使用.list文件里的绝对全路径。"),
                            scale=10,
                        )

                    gr.Markdown(value=i18n("1Aa-文本内容"))
                    with gr.Row():
                        gpu_numbers_text = gr.Textbox(label=i18n("GPU卡号以-分割，每个卡号一个进程"), value=f"{GPUS_STR}-{GPUS_STR}", interactive=True)
                        bert_pretrained_dir = gr.Textbox(label=i18n("预训练的中文BERT模型路径"), value="GPT_SoVITS/pretrained_models/chinese-roberta-wwm-ext-large", interactive=False, lines=2)
                        button1a_open = gr.Button(i18n("开启文本获取"), variant="primary", visible=True)
                        button1a_close = gr.Button(i18n("终止文本获取进程"), variant="primary", visible=False)
                        info1a = gr.Textbox(label=i18n("文本进程输出信息"))

                    gr.Markdown(value=i18n("1Ab-SSL自监督特征提取"))
                    with gr.Row():
                        gpu_numbers_ssl = gr.Textbox(label=i18n("GPU卡号以-分割，每个卡号一个进程"), value=f"{GPUS_STR}-{GPUS_STR}", interactive=True)
                        ssl_pretrained_dir = gr.Textbox(label=i18n("预训练的SSL模型路径"), value="GPT_SoVITS/pretrained_models/chinese-hubert-base", interactive=False, lines=2)
                        button1b_open = gr.Button(i18n("开启SSL提取"), variant="primary", visible=True)
                        button1b_close = gr.Button(i18n("终止SSL提取进程"), variant="primary", visible=False)
                        info1b = gr.Textbox(label=i18n("SSL进程输出信息"))

                    gr.Markdown(value=i18n("1Ac-语义token提取"))
                    with gr.Row():
                        gpu_numbers_sem = gr.Textbox(label=i18n("GPU卡号以-分割，每个卡号一个进程"), value=f"{GPUS_STR}-{GPUS_STR}", interactive=True)
                        pretrained_s2G_ = gr.Textbox(label=i18n("预训练的SoVITS-G模型路径"), value=pretrained_sovits_name[-int(version[-1]) + 2], interactive=False, lines=2)
                        button1c_open = gr.Button(i18n("开启语义token提取"), variant="primary", visible=True)
                        button1c_close = gr.Button(i18n("终止语义token提取进程"), variant="primary", visible=False)
                        info1c = gr.Textbox(label=i18n("语义token提取进程输出信息"))

                    gr.Markdown(value=i18n("1Aabc-训练集格式化一键三连"))
                    with gr.Row():
                        button1abc_open = gr.Button(i18n("开启一键三连"), variant="primary", visible=True)
                        button1abc_close = gr.Button(i18n("终止一键三连"), variant="primary", visible=False)
                        info1abc = gr.Textbox(label=i18n("一键三连进程输出信息"))

                    pretrained_s2G.change(sync, [pretrained_s2G], [pretrained_s2G_])

                    button1a_open.click(open1a, [inp_text, inp_wav_dir, exp_name, gpu_numbers_text, bert_pretrained_dir], [info1a, button1a_open, button1a_close])
                    button1a_close.click(close1a, [], [info1a, button1a_open, button1a_close])

                    button1b_open.click(open1b, [inp_text, inp_wav_dir, exp_name, gpu_numbers_ssl, ssl_pretrained_dir], [info1b, button1b_open, button1b_close])
                    button1b_close.click(close1b, [], [info1b, button1b_open, button1b_close])

                    button1c_open.click(open1c, [inp_text, exp_name, gpu_numbers_sem, pretrained_s2G], [info1c, button1c_open, button1c_close])
                    button1c_close.click(close1c, [], [info1c, button1c_open, button1c_close])

                    button1abc_open.click(open1abc, [inp_text, inp_wav_dir, exp_name, gpu_numbers_text, gpu_numbers_ssl, gpu_numbers_sem, bert_pretrained_dir, ssl_pretrained_dir, pretrained_s2G], [info1abc, button1abc_open, button1abc_close])
                    button1abc_close.click(close1abc, [], [info1abc, button1abc_open, button1abc_close])

                with gr.TabItem(i18n("1B-微调训练")):
                    gr.Markdown(value=i18n("1Ba-SoVITS训练。用于分享的模型文件输出在SoVITS_weights下。"))
                    with gr.Row():
                        batch_size = gr.Slider(minimum=1, maximum=40, step=1, label=i18n("每张显卡的batch_size"), value=DEFAULT_BATCH_SIZE, interactive=True)
                        total_epoch = gr.Slider(minimum=1, maximum=25, step=1, label=i18n("总训练轮数total_epoch，不建议太高"), value=8, interactive=True)
                        text_low_lr_rate = gr.Slider(minimum=0.2, maximum=0.6, step=0.05, label=i18n("文本模块学习率权重"), value=0.4, interactive=True)
                        save_every_epoch = gr.Slider(minimum=1, maximum=25, step=1, label=i18n("保存频率save_every_epoch"), value=4, interactive=True)

                    with gr.Row():
                        if_save_latest = gr.Checkbox(label=i18n("是否仅保存最新的ckpt文件以节省硬盘空间"), value=True, interactive=True)
                        if_save_every_weights = gr.Checkbox(label=i18n("是否在每次保存时间点将最终小模型保存至weights文件夹"), value=True, interactive=True)
                        gpu_numbers_sovits_train = gr.Textbox(label=i18n("GPU卡号以-分割，每个卡号一个进程"), value=f"{GPUS_STR}", interactive=True)

                    with gr.Row():
                        button1Ba_open = gr.Button(i18n("开启SoVITS训练"), variant="primary", visible=True)
                        button1Ba_close = gr.Button(i18n("终止SoVITS训练"), variant="primary", visible=False)
                        info1Ba = gr.Textbox(label=i18n("SoVITS训练进程输出信息"))

                    gr.Markdown(value=i18n("1Bb-GPT训练。用于分享的模型文件输出在GPT_weights下。"))
                    with gr.Row():
                        batch_size_gpt = gr.Slider(minimum=1, maximum=40, step=1, label=i18n("每张显卡的batch_size"), value=DEFAULT_BATCH_SIZE, interactive=True)
                        total_epoch_gpt = gr.Slider(minimum=2, maximum=50, step=1, label=i18n("总训练轮数total_epoch"), value=15, interactive=True)
                        save_every_epoch_gpt = gr.Slider(minimum=1, maximum=50, step=1, label=i18n("保存频率save_every_epoch"), value=5, interactive=True)
                        if_dpo = gr.Checkbox(label=i18n("是否开启dpo训练选项(实验性)"), value=False, interactive=True)

                    with gr.Row():
                        if_save_latest_gpt = gr.Checkbox(label=i18n("是否仅保存最新的ckpt文件以节省硬盘空间"), value=True, interactive=True)
                        if_save_every_weights_gpt = gr.Checkbox(label=i18n("是否在每次保存时间点将最终小模型保存至weights文件夹"), value=True, interactive=True)
                        gpu_numbers_gpt_train = gr.Textbox(label=i18n("GPU卡号以-分割，每个卡号一个进程"), value=f"{GPUS_STR}", interactive=True)

                    with gr.Row():
                        button1Bb_open = gr.Button(i18n("开启GPT训练"), variant="primary", visible=True)
                        button1Bb_close = gr.Button(i18n("终止GPT训练"), variant="primary", visible=False)
                        info1Bb = gr.Textbox(label=i18n("GPT训练进程输出信息"))

                    button1Ba_open.click(open1Ba, [batch_size, total_epoch, exp_name, text_low_lr_rate, if_save_latest, if_save_every_weights, save_every_epoch, gpu_numbers_sovits_train, pretrained_s2G, pretrained_s2D], [info1Ba, button1Ba_open, button1Ba_close])
                    button1Ba_close.click(close1Ba, [], [info1Ba, button1Ba_open, button1Ba_close])

                    button1Bb_open.click(open1Bb, [batch_size_gpt, total_epoch_gpt, exp_name, if_dpo, if_save_latest_gpt, if_save_every_weights_gpt, save_every_epoch_gpt, gpu_numbers_gpt_train, pretrained_s1], [info1Bb, button1Bb_open, button1Bb_close])
                    button1Bb_close.click(close1Bb, [], [info1Bb, button1Bb_open, button1Bb_close])

                with gr.TabItem(i18n("1C-推理")):
                    SoVITS_names, GPT_names = get_weights_names()
                    gr.Markdown(value=i18n("选择训练完存放在SoVITS_weights和GPT_weights下的模型。默认的一个是底模，体验5秒Zero Shot TTS用。"))
                    with gr.Row():
                        GPT_dropdown = gr.Dropdown(label=i18n("*GPT模型列表"), choices=sorted(GPT_names, key=custom_sort_key), value=pretrained_gpt_name[0], interactive=True)
                        SoVITS_dropdown = gr.Dropdown(label=i18n("*SoVITS模型列表"), choices=sorted(SoVITS_names, key=custom_sort_key), value=pretrained_sovits_name[0], interactive=True)

                    with gr.Row():
                        gpu_number_infer = gr.Textbox(label=i18n("GPU卡号,只能填1个整数"), value=GPUS_STR, interactive=True)
                        refresh_button = gr.Button(i18n("刷新模型路径"), variant="primary")
                    refresh_button.click(fn=change_choices, inputs=[], outputs=[SoVITS_dropdown, GPT_dropdown])

                    with gr.Row():
                        batched_infer_enabled = gr.Checkbox(label=i18n("启用并行推理版本(推理速度更快)"), value=False, interactive=True)

                    with gr.Row():
                        open_tts = gr.Button(value=i18n("开启TTS推理WebUI"), variant="primary", visible=True)
                        close_tts = gr.Button(value=i18n("关闭TTS推理WebUI"), variant="primary", visible=False)
                        tts_info = gr.Textbox(label=i18n("TTS推理WebUI进程输出信息"))

                    open_tts.click(change_tts_inference, [bert_pretrained_dir, ssl_pretrained_dir, gpu_number_infer, GPT_dropdown, SoVITS_dropdown, batched_infer_enabled], [tts_info, open_tts, close_tts])
                    close_tts.click(change_tts_inference, [bert_pretrained_dir, ssl_pretrained_dir, gpu_number_infer, GPT_dropdown, SoVITS_dropdown, batched_infer_enabled], [tts_info, open_tts, close_tts])

                version_checkbox.change(switch_version, [version_checkbox], [pretrained_s2G, pretrained_s2D, pretrained_s1, GPT_dropdown, SoVITS_dropdown])

            with gr.TabItem(i18n("2-GPT-SoVITS-变声")):
                gr.Markdown(value=i18n("施工中，请静候佳音"))

        app.queue(max_size=64).launch(
            server_name="0.0.0.0",
            inbrowser=True,
            share=is_share,
            server_port=webui_port_main,
            quiet=True,
            max_threads=32,  # 필요하면 16~64 사이로 조절
        )


if __name__ == "__main__":
    main()